Re: TMD Partial Fitting Problem

From: Gumbart, JC (gumbart_at_physics.gatech.edu)
Date: Tue Jun 11 2019 - 18:53:35 CDT

Did you keep in mind this aspect of TMD?

after first aligning the target structure to the current coordinates

It can be pretty difficult to keep in mind all of the different TMD options at once. When in doubt, simplify what you are trying to do and see that it behaves as expected first.

Best,
JC

On Jun 11, 2019, at 8:26 AM, mert gölcük <mertgolcuk_at_gmail.com<mailto:mertgolcuk_at_gmail.com>> wrote:

Dear all,
What we want to do is to use a different set of atoms for fitted atoms and biased atoms. As it is explained here
https://www.ks.uiuc.edu/Research/namd/2.10b1/ug/node47.html
We set the altloc field for the fitted atoms to '1'.

The system has 186 residues. 125 alpha carbon atoms were fitted and 61 were biased. When we run 50000 steps with 2 fs time step. The TMD spring constant was 200 kcal/mol*A^2 and last step of the TMD was set 100000 (i.e. the end of the simulation). Interestingly, The RMSD of the TMD atoms to the target in the log file started with 0.115181 A and went to 0 in just 2000 steps. This issue happens in both NAMD-2.13-CUDA and NAMD-2.13-Multicore versions.

Here we have to point out that the biased atoms are located in a rather rigid part of the protein. In other words, if the biased 61 atoms are aligned with their target directly then the RMSD is only 0.0005 A. What we aim to observe is the TMD is the movement of this region of the protein with respect to the remaining part.

From what we see is that in our simulations the 125 atoms that are supposed to be fitted are apparently not fitted. Or the TMD forces are not applied in a sense so that it causes the region to move with respect to the remaining of the protein. One solution to this might be to constrain the fitted atoms, but I am more inclined to think that this shouldn’t be required in the TMD simulations.

Any comments or suggestions would be greatly appreciated.

Initial RMSD values of the system were measured 1.54 Å for whole system, 6.06 Å for the whole protein, 19.21 Å for whole TMD domain and 33.54 Å for biased atoms. After the simulation started, we were obtained 0.0113 and 0.115181 values from the log file and this values show that targetRMSD and currentRMSD values respectively.

Thank you all.
--
Mert GÖLCÜK

This archive was generated by hypermail 2.1.6 : Wed Dec 04 2019 - 23:20:50 CST